In this paper, a method based on modified incremental conductance with intelligence predication was proposed as an application on photovoltaic module array (PVMA) for the purpose of maximum power point tracking (MPPT), which served to improve the power generation performance of PVMA. Since the PVMA is affected by solar irradiance and temperature change, the output maximum power point (MPP) changes accordingly. In addition, owing to the fixed length of tracking pace for the conventional incremental conductance method (ICM), the time required for tracking to the MPP may take longer. Therefore, the modified ICM with step size predication was proposed in this paper. The length of tracking pace could automatically adapt to the gradient of the power–voltage characteristic curve for PVMA. The 0.8 times voltage for the MPP of PVMA under standard test conditions would serve as the initial voltage for MPPT, which would be utilized to increase the output power of PVMA. In this study, the 62050H-600S programmable DC power supply produced by Chroma ATE Inc. was applied to simulate the output characteristics of 4-series-3-parallel PVMA. In contrast, the conventional and modified ICMs were applied for MPPT purposes. The actual test results proved that the modified ICM proposed in this paper provided a better response in tracking speed and steady-state performance than the conventional ICM under different solar conditions.